Prediction of internal migration after an earthquake with call detail records

The 2023 Kahramanmaraş Earthquake resulted in over 55,000 deaths. Efficient evacuation management can significantly reduce secondary casualties and optimize resource allocation. This study explores the application of Call Detail Records (CDR) in times of crisis, with a particular focus on the consequences of earthquakes. We focus on two issues: the post-earthquake evacuation behavior and prediction of population movements after earthquakes. We use machine learning models with distance-normalized features inspired by the Gravity model to predict population movements immediately after earthquake. The experiments show that our model can predict evacuation flows between different districts. Our main findings are that population distribution and earthquake intensity are the primary factors of evacuation patterns. Comparative analysis between Turkish and Syrian populations shows the same feature importance rankings, but distinct pattern distributions.